Subject: Re: [vserver] vserver scalability
From: Eugen Leitl <eugen@leitl.org>
Date: Tue, 15 Jun 2010 11:54:07 +0200

On Mon, Jun 14, 2010 at 11:14:26PM +0200, Andre Timmermann wrote:
> Hi,
> 
> Am Donnerstag, den 03.06.2010, 21:32 +0200 schrieb Herbert Poetzl:
> 
> > > Am particularly interested to know how many virtual servers 
> > > can be spawned in a single host?
> > 
> > we stopped testing around 40.000 guests on a single
> > host, because it became somwhat slow and the 64k
> > process limit was an issue back then ....
> 
> I am pretty sure that it is POSSIBLE to start several hundrets or even
> thousends of guests on one physical machine, but the real question is
> how many guests are REASONABLE on on physical machine.
> 
> In my experience the main impact is I/O on normal server hardware (e.g
> Dual Intel Xeon CPU / 8 cores / 12 GB RAM / two disks with raid 1) 

Something which can be adressed with SSDs.
 
> If each instance operates an own mysql-server, then it gets worse.
> 
> So the second question should be: what will be done/installed on the
> guests?
> 
> Greetz,
> Andre
> 
-- 
Eugen* Leitl <a href="leitlhttp://leitl.org">leitl</a> http://leitl.org
______________________________________________________________
ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A  7779 75B0 2443 8B29 F6BE